Structural and Functional Illumination of the Human Dark Proteome (Trilogy III — Oncology Edition): Chimeric Pyrin/Death-Fold–G3BP1 Stress-Granule Architecture in C3orf38 (Q5JPI3 / DUF4518), Membrane-Tethered GH65/PGGHG Collagen-Like Glycosidase Fold in KIAA2013 (Q8IYS2 / DUF2152), and a 143-Residue Pfam Boundary Error & 14-3-3 Regulatory Latch in Melanoma/Cachexia Driver KIAA0930 (Q6ICG6 / DUF2045)
Year: 2026
Type: report
Abstract: Computational structural and oncology-focused functional characterization of three human 'Dark Proteome' proteins carrying Domains of Unknown Function (DUFs): (1) C3orf38 (UniProt Q5JPI3 / Pfam PF29669 DUF4518_N & PF15008 DUF4518), associated with prostate and esophageal carcinoma and binding the anti-apoptotic protein BAG4 (SODD), is resolved as a chimeric two-domain adapter uniting an N-terminal 6-helix antiparallel Pyrin / Death-like bundle (Phe8-Leu105, CATH 1.10.533.10, 2.62 A Core C-alpha RMSD vs human ASC2 PYD PDB 2HM2) with a C-terminal G3BP1 / NTF2-like stress-granule cone domain (Phe138-Gly277, CATH 3.10.450.50, Foldseek Prob = 1.000, 2.15 A Core C-alpha RMSD across 96/111 atoms, TM-score = 0.728 vs human G3BP1 PDB 6TA7); (2) KIAA2013 (UniProt Q8IYS2 / Pfam PF10222 DUF2152), associated with acute lymphoblastic leukemia, ovarian/pancreatic/endometrial cancers, and Alzheimer disease, is identified with Foldseek Prob = 1.000 (E = 4.35e-14) as a membrane-tethered structural paralog of human PGGHG (Protein-glucosylgalactosylhydroxylysine glucosidase / GH65 family, CATH 2.70.98.50; 3.52 A Core RMSD & TM-score = 0.403 vs GH65 alpha-1,2-glucosidase PDB 7FE3), conserving 3 disulfide bridges and 7 acidic active-site pocket residues; and (3) KIAA0930 (UniProt Q6ICG6 / C22orf9 / Pfam PF09741 DUF2045), strongly linked in Open Targets to cutaneous melanoma (0.752), basal/squamous cell carcinoma, lung cancer, and cancer cachexia, is shown to fold into a 15-stranded C2/Ig-like beta-sandwich (CATH 2.60.40.150, Prob = 0.984) interrupted after residue 260 by a 114-residue hyper-phosphorylated 14-3-3-binding loop (261-374, housing all 10 pSer/pThr sites) before returning at the C-terminus (Leu375-Leu404, pLDDT = 96.0) to form Strand beta-15 and Helix alpha-6, revealing a 143-residue C-terminal boundary error in Pfam PF09741 (annotated as 30-261). [Open Science v2.0 5-Part Multi-Modal Publication Suite: Includes (1) 2400x1800 4-panel composite publication figures combining 1D Pfam vs. v2.0 domain tracks & pLDDT profiles, 2D AlphaFold PAE matrices, 3D DSSP secondary-structure cartoons & crystal superpositions, and atomic active-site/salt-bridge callouts in Angstroms; (2) pre-aligned 3D superposition PDBs & turnkey PyMOL .pml session scripts; (3) cross-species vertebrate/bilaterian ortholog MSAs (.fasta); (4) gnomAD/ClinVar/GTEx/HPA genetics & expression summaries (.json); and (5) ready-to-paste EMBL-EBI Pfam DESC & UniProt Community Curation files.]
Author: Chappell, Corey
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7224135271